|
- mise à jour
du
- 17 juillet
2022
- Anim
Cogn
- 2023 Mar
16
-
|
- Contagious
yawning in African painted dogs
- Kanako Ake, Nobuyuki Kutsukake
-
- Department of Evolutionary
Studies of Biosystems, Sokendai
- Hayama, Kanagawa,
Japan
|
|
- Tous
les articles sur la contagion du
bâillement
- All
articles about contagious
yawning
-
- Abstract
- Contagious yawning (CY), which is yawning
elicited by sensing another yawning, has been
observed only in social species and is
considered linked to high sociality. Although
this idea&emdash;the social communication
hypothesis&emdash;is supported by pre- vious
studies, investigating the occurrence of CY in
various species remains necessary.The authors
investigated the occurrence of CY in one of the
most social canine species, the African painted
dog (Lycaon pictus). We recorded 1387 yawn
events from five pairs (10 individuals) in
captivity. Temporal analysis showed that
subsequent yawns occurred frequently within 15 s
or 30 s after spontaneous yawns (SYs). SYs that
were detectable by another individual (i.e.,
visible to the other individual or performed in
close proximity) were more likely to elicit
subsequent yawns by the other individual.
However, the influence of contextual factors on
CY differed according to the time window,
implying that a conservative time window should
be used to reduce overcounting of CY and
misattribution of its determinants. The
proportion of CY was positively related to time
spent in physical proximity to another, a proxy
for a social bond. Overall, our results provide
the first evidence of CY in African painted dogs
and further support the notion that such
behavior is prevalent among social animals. They
also strongly imply that an appropriate time
window should be used to define yawn
contagion.
-
- Résumé
- La contagion du bâillement,
c'est-à-dire le bâillement
provoqué par la perception d'un autre
bâillement, n'est observé que chez
les espèces ayant une vie sociale. Elle
et est considérés comme
liée à une grande
sociabilité. Bien que cette idée -
l'hypothèse de la communication sociale -
soit étayée par des études
antérieures, il reste nécessaire
d'étudier l'occurrence de la contagion du
bâillement chez diverses espèces.
Les auteurs ont étudié
l'occurrence de cette contagion au sein d'une
des espèces canines à vie sociale
évoluée, le chien Lycaon pictus.
Ils ont enregistré 1387 bâillements
chez cinq paires d'individus en
captivité. L'analyse temporelle a
montré que les bâillements
déclenchés se produisaient
fréquemment dans les 15 s ou 30 s
après les bâillements
spontanés (émetteur). Les
bâillements spontanés
détectables par un autre individu
(c'est-à-dire visibles par l'autre
individu ou effectués à
proximité) étaient plus
susceptibles de provoquer des bâillements
ultérieurs chez l'autre individu.
Cependant, l'influence des facteurs contextuels
sur cette contagion diffère selon la
temporalité, ce qui implique
d'éviter l'attribution erronée de
ces bâillements (synchronisation des
activités identiques liée aux
rythmes circadiens et sans lien avec la
réelle contagion).
-
-
- Yawning is of interest as an evolutionarily
conserved motor action observed across diverse
taxa. Numerous functional hypotheses have been
proposed for yawning. Some of these hypotheses
focus on physiological functions: absorbing much
oxygen into the body (the respiratory
hypothesis); stimulating the body under
exhausted or sleepy states (the stress-related
arousal hypothesis); and cooling the brain by
absorbing cool air (the brain-cooling
hypothesis). The social communication
hypothesis, on the other hand, states that
yawning functions as a social cue or signal to
express physiological or emotional status to
other individuals (Guggisberg et al. 2010;
Gallup 2022).
-
- The social communication hypothesis has been
supported by the phenomenon that sensing other
individuals yawning often elicits yawning in
humans (Krestel et al. 2018). Such contagious
yawning (CY) is categorized into automatic
mimicry or emotional contagion, which is both
considered to promote synchronization among
group members (Palagi et al. 2020). Thus, the
social communication hypothesis predicts that
animals with a high degree of cooperative inter-
action among group members will show CY (Palagi
et al. 2020; Norscia et al. 2021). Behavioral
studies have tested this prediction and found
that group-living animals show CY (Gallup 2022).
Moreover, the social communication hypothesis
predicts that CY promotes behavioral
synchronization within a group, which has been
observed in lions (Panthera leo, Casetta et al.
2021). The hypothesis further predicts a
positive relationship between the frequency of
CY and relationship closeness. Indeed, yawning
is more contagious between bonded individuals
than between weakly bonded individuals in humans
(Homo sapiens, Norscia et al. 2020), bonobos
(Pan paniscus, Palagi et al. 2014), chimpanzees
(Pan troglodytes, Demuru and Palagi 2012;
Campbell and de Waal 2011), wolves (Canis lupus
lupus, Romero et al. 2014), and domestic pigs
(Sus scrofa, Norscia et al. 2021).
-
- Several studies, however, failed to
demonstrate the relation- ship between the
frequency of CY and relationship closeness
(e.g., budgerigars, Melopsittacus undulatus,
Gallup et al. 2015; see review by Massen and
Gallup 2017). Although empirical studies of the
occurrence and determinants of CY have
accumulated, at least two points require further
examination. First, the phylogenetic
distribution of CY remains unclear, as
systematic research on diverse species is
limited (Gallup 2010; Guggisberg et al. 2010).
Thus, research on the occurrence of CY in
diverse taxa is needed. Second, a methodological
controversy exists regarding the definition of
CY. The prevalent method of determining CY
involves evaluating the frequency of yawning
within a defined temporal interval, with and
without prior exposure to yawning stimuli from
another individual. The majority of these
studies have defined CY as all following yawns
within 3&endash;5 min after the detection of a
yawn by another individual. However, this time
window has been noted to be too long for
determining whether the following yawn is
spontaneous or contagious (Massen and Gallup
2017).
-
- For example, in bonobos, the occurrence of
CY is detected only early in the time window
(i.e., the first minute), not in the second or
third minute (Norscia et al. 2022). As such, the
use of an arbitrary time window can result in an
inaccurate determination of CY. To date, no
studies have examined the impacts of different
time windows on biological conclusions, such as
attribution of the determinants of CY. If the
peak time window in which a response yawn
follows a preceding spontaneous yawn (SY) is
less than 3 min in the studied species,
overcounting of CY due to the inclusion of SY or
misattribution of the contextual or social
factors influencing CY could occur. Other
methodologies have been used for CY studies
(e.g., Romero et al. 2014; Miller et al. 2012;
Massen et al. 2016; Gallup et al. 2022). Some
stud- ies showed video clips of yawning to
subject animals (e.g., Amici et al. 2014; Reddy
et al. 2016; Pedruzzi et al. 2022).
-
- Here, we conducted behavioral observations
of yawning in one of the most social canine
species, the African painted dog (Robbins 2000;
Walker et al. 2017). This cooperatively breeding
carnivore lives in a cohesive group of
3&endash;20 members per pack. This species
exhibits various social behaviors, including
synchronized actions (Walker et al. 2017),
cooperative hunting, and food sharing
(regurgitation to pups or weakened adult
individuals, Creel and Creel 1995). They have
the largest repertoire of vocal communication
among canids, which depends on social context
(Robbins 2000). Based on the cooperation scores
proposed by Smith et al. (2012), canine species
show uniquely strong cooperative behavior, with
the highest score for cooperation in African
painted dogs. Thus, this species is an ideal
subject for studying the occurrence of CY.
-
- Previous studies using the time window
method have confirmed that the peak time window
in which CY occurs is clearly shorter than 3 min
(Table 1). Therefore, we first examined whether
successive yawns by two individuals occurred
within a relatively short time window. To this
end, we examined the peak time window in which
CY occurred frequently to determine an
appropriate time window for operationally
defining CY in this species. To assess the
occurrence of CY, we tested whether yawning is
contagious when an individual can detect another
individual yawning. If this species exhibits CY,
individuals will yawn more frequently after
visible SYs than after nonvisible yawns
(Prediction 1a), as well as when two individuals
are in close proximity than at a distance
(Prediction 1b). Next, we predicted that
different contextual or social factors would
affect the occurrence of CY depending on the
time window used for the operational definition
(Prediction 2). We tested this through a
comparison of the results obtained using
different time windows. Furthermore, as bonded
individuals tend to detect the actions of others
(Massen and Gallup 2017), yawning should be
detected by bonded individuals, resulting in a
high frequency of CY in such individuals.
Therefore, we predicted that strongly bonded
pairs would show CY more frequently than weakly
bonded pairs (Prediction 3).
-
 -
- Discussion
-
- We found that yawn contagion occurred
immediately after visible SYs, with steep and
gradual peak windows observed within
approximately 15 and 30 s, respectively). Model
1 showed that responders yawned more frequently
after visible SYs than after nonvisible SYs,
supporting Pre- diction 1a (Table 3). The
positive effect of visibility of SYs on FYs is a
key factor used to define the occurrence of CY
in previous studies (e.g., Gallo et al. 2021;
Casetta et al. 2021). Model 1 revealed that
physical proximity between the trigger and
responder during an SY positively affected the
occurrence of FYs (Table 3). This result
supports Prediction 1b. A similar effect of
physical proximity on CY has been reported in
previous studies (e.g., Norscia et al. 2021). As
African painted dogs show less facial expression
than other social canine species such as wolves
(Kleiman 1967; Fox 1971), they may require close
proximity to detect a trigger. However, once
they detect others yawn, one yawn is sufficient
to elicit CY, as indicated by our finding that
the number of SYs did not affect the occurrence
of CY. The effect of the detectability of SYs
and our observation of FY occurrence immediately
after SYs imply that African painted dogs
exhibit CY. Other social canine species also
exhibited CY (wolves: Romero et al. 2014). While
some studies showed CY of domestic dogs (Canis
lupus familiaris) in response to human yawning
(e.g., Joly-Mascheroni et al. 2008), the results
are controversial (e.g., Buttner and Strasser
2014; Harr et al. 2009). In addition, CY in
response to conspecifics has not yet been
demonstrated (Harr et al. 2009).
-
- Our analysis implies that the selection of
an appropriate time window is critical. Using a
3 min time window to define CY could lead to
overcounting yawns as CYs, as proposed
previously (Massen and Gallup 2017). We should
note that some yawns occurring in the 15 s and
30 s time windows could also be incorrectly
categorized as CY. How- ever, the frequency of
the misidentification of SY as CY in the 3 min
window led to notable overcounting. Some studies
have shown that the peak CY occurrence times are
within 1 min or less; nonetheless, arbitrary
time windows are commonly used to define an FY
after a trigger yawn as a potential CY (Table
1). Campbell and Cox (2019) showed that the 3
min time window is suitable for chimpanzees, but
noted that the same result would not be expected
for other species. To minimize overcounting and
misevaluation, time windows for FYs shorter than
3 min have been used in some studies (i.e.,
Demuru and Palagi 2012; Wojczulanis-Jakubas et
al. 2019; Miller et al. 2012). If CY is truly a
type of motor mimicry, CY would be predicted to
occur within about 5 s after an SY, in
accordance with other motor mimicry behaviors
(Prochazkova and Kret 2017). In this study,
frequent FYs indeed occurred between 3 and 8 s
after SYs (Fig. 2). Most previous studies,
however, did not subdivide the observation
window into shorter periods of 30 s less
(however, 20 s was arbitrarily selected by
Wojczulanis-Jakubas et al. 2019). Subdivision of
the time window and consideration of
species-specific peak time windows are essential
to avoiding missing CY or misattributing an SY
as CY.
-
- Model 1 indicated that the difference among
time win- dows affected the results for
significant explanatory variables, which
supported Prediction 2. We found a negative
effect of distance between the trigger and
responder during an SY on the proportion of FYs
when using the 15 and 30 s time windows but not
the 3 min time window. Similarly, different
effects on CY frequency were found with
different time windows (Table 4), as reported in
previous studies. In humans and bonobos, Palagi
et al. (2014) defined all FYs within 3 min after
an SY as CY. Their study showed that frequent CY
occurs among strongly bonded individuals in the
first minute, whereas CY among weakly bonded
individuals is more frequent in the second and
third minute. Based on this result, they
suggested that the latency of CY differs
according to familiarity. In wild gelada
baboons, CY was confirmed when the analysis
employed a time window including the peak time
(the second minute) but not when the analysis
considered only the first minute (Gallo et al.
2021). The results of these studies raise the
possibility that some reported CY actually
included SYs due to the use of too-long a time
window, indicating that species-specific time
windows should be used to define CY and evaluate
the fac- tors affecting CY.
-
- Hereafter, we discuss testing predictions
using the results for FYs within 30 s, as the
results did not differ between the 15 s and 30 s
time windows. In agreement with Prediction 3,
the proportion of time spent in physical
proximity per day positively affected the
proportion of CY per individual per day.
Similarly, previous studies have shown that CY
occurs more frequently in more strongly bonded
pairs of individuals than in weakly bonded pairs
as measured based on affilia- tive behavior
(Demuru and Palagi 2012; Palagi et al. 2014;
Gallo et al. 2021; Norscia et al. 2021; but see
Gallup et al. 2015). All pairs of individuals in
this study were related and bonded to the extent
that they could live in a limited space with no
severe fights, as the pairs were selected by zoo
staff based on observed social interactions. Our
scan sam- pling implied that paired individuals
spent approximately half of the time within 1 BL
of each other (mean = 0.44, standard deviation =
0.12), which allowed them to detect the SYs.
Such same-sexed relatives with strong social
bonds can be found in the wild, in which related
same-sex siblings disperse and form a pack by
encountering an opposite-sex group (Creel and
Creel 2002). They maintain close distances and
frequently interact with group members to
synchronize their behaviors, using cues such as
sneezing before departure for hunting (Walker et
al. 2017). Strongly bonded pairs are more likely
to detect facial expressions than weakly bonded
pairs (i.e., attention bias; Massen and Gallup
2017). There- fore, it is plausible that the
stronger the bond of a pair is, the more
reliably they can detect SYs from each other,
resulting in more frequent CY.
-
- In summary, this is the first study
demonstrating CY in the African painted dog, one
of the most social canine species. Our results
have at least two general implications. First,
the occurrence of CY in this species is in line
with the idea that social species exhibit CY.
African painted dogs show various social
behaviors that require coordination with other
pack members (e.g., cooperative hunting). The
social communication hypothesis predicts that CY
occurs in species with a high degree of
cooperative interaction among conspecifics
(Guggisberg et al. 2010). In addition, previous
studies have shown that species with few
cooperative interactions between individuals do
not exhibit CY, such as gorillas (Amici et al.
2014; Palagi and Norscia 2019) and red-footed
tortoises (Wilkinson et al. 2011, Table 1).
However, some studies of CY on social species
have failed to show CY due to rare occurrences
of yawning during observations (e.g., ravens,
Corvus corax: Gallup et al. 2022; common
marmosets, Callithrix jacchus: Massen et al.
2016). Accumulating data regarding CY in both
social and nonsocial species will be required to
test the social communication hypothesis. As a
second general implication, our results show
that the most commonly used arbitrary time
window was not appropriate for detecting CY in
this species. Using an appropriate time window
based on the peak timing of CY in each species,
rather than a conventionally used but arbitrary
time window that may not fit the study species,
is essential to evaluate the presence of CY and
the contextual or social factors affecting CY
frequency.
|
|